1885.] On the transpiration-stream in cut branches. 3.59 



If the branch has been previously " double-sawn " it will not 

 (at least at first) be uniformly coloured, but will show two some- 

 what wedge-shaped patches of colour each of which corresponding 

 in position to one of the lines of colour in figs. 3 and 4. These are 

 shown in fig. 5. It will be seen that one side of the wedge verti- 

 cally below the inner edge of the lower saw-cut is sharply defined, 

 while the other edge is feathered. This shows that the suction 

 is strongest where the stream passes by the edge of the saw- 

 cut, and gets weaker along radii further removed from this point. 



The chief point of interest connected with this subject is that 

 similar appearances can be produced by forcing eosin solution 

 through doubly sawn-branches ; it is impossible to distinguish 

 such a specimen from a transpiring branch which has been doubly 

 sawn and placed in eosin. The wedge-shaped patches of colour, 

 with one feathery edge, as seen in fig. 5, may also be produced 

 by attaching a double sawn branch to a pump while the other end 

 is in a mixture of carmine and water. 



§ 8. Conclusion on Dufour's Experiments. 



We have not enough data to criticise Dufour's paper. Elfving* 

 and Scheit"f* have pointed out a possible source of error from the 

 effect of air on the cut ends of the branch. We have shown the 

 importance of great care in making the saw-cuts ; it is easy (even 

 without carelessness) to leave a small region of transverse section 

 unsevercd, and we have shown how readily a current may be 

 conveyed over a very narrow bridge of Avood. Both these sources 

 of error may have contributed, in different ways, to leading Dufour 

 to an untenable result. That it is untenable the following re- 

 capitulation of our results seem to prove. 



I. Double sawing does produce a great depression in the rate 

 of absorption. No one who will make a single careful potometer 

 experiment (with an Angiosperm) will be able to doubt it for a 

 moment. 



II. Double sawing does not absolutely interrupt the filtration 

 current J. 



[Absolute stoppage of the current is required for Dufour's 

 argument, since the withering experiments which are coordinated 

 with the pressure experiments in the argument give no quantitative 

 record of the amounts of water passing by the saw-cuts.] 



III. Double sawing produces a depression in the filtration 

 current similar to, though possibly greater than, the depression 

 caused in the transpiration current. 



* Bot. Zeitung, 1884, No. 13. t Acta Soc. Fennica, t. xiv. 1884. 



J Scheit has shown this fact, loc. cit. 



